Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <19415767029b479a508378abaf8c6c5a1d048add.camel@xry111.site>
Subject: Re: [PATCH] LoongArch: mm: Refactor TLB handlers
From:   Xi Ruoyao <xry111@xry111.site>
To:     Huacai Chen <chenhuacai@kernel.org>, Rui Wang <wangrui@loongson.cn>
Cc:     loongarch@lists.linux.dev, WANG Xuerui <kernel@xen0n.name>,
        linux-kernel@vger.kernel.org
Date:   Sun, 09 Oct 2022 18:37:28 +0800
In-Reply-To: <CAAhV-H6e9ZyC8TkpWK3Etgz4-kEEEU4Hj=i5obMFk=URgCwrgw@mail.gmail.com>
References: <20221004140230.748788-1-wangrui@loongson.cn>
         <CAAhV-H6e9ZyC8TkpWK3Etgz4-kEEEU4Hj=i5obMFk=URgCwrgw@mail.gmail.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.46.0 
MIME-Version: 1.0
Precedence: bulk

On Sun, 2022-10-09 at 14:13 +0800, Huacai Chen wrote:
> Queued for loongarch-next, thanks.

> And fixes the concurrent modification issue of fast path for huge
> pages.
> This issue will occur in the following steps:
>=20
>    CPU-1 (In TLB exception)         CPU-2 (In THP)
> 1: Load PMD entry (HUGE=3D1)
> 2: Goto huge path
> 3:                                  Store PMD entry (HUGE=3D0)
> 4: Reload PMD entry (HUGE=3D0)
> 5: Fill TLB entry (PA is incorrect)

Is this issue causing problems in practice?  If true should we pick this
for loongarch-fixes and backport into 5.19 & 6.0 too?

--=20
Xi Ruoyao <xry111@xry111.site>
School of Aerospace Science and Technology, Xidian University